In vitro protein synthesis using natural messenger RNAs requires the presence of initiation factor IF3. When synthetic single-stranded messenger RNAs are used, IF3 is not essential, but stimulates the reaction. These observations apply to in vitro systems from mammals, plants, inverebrates, and single-celled organisms. In a preliminary study of the function of E. coli IF3, I found that it bound AUG, the protein synthesis initiation codon, quite strongly on a per nucleotide basis, and that it nonspecifically bound other oligonucleotides weakly on a per nucleotide basis. I propose to measure the binding of E. coli IF3 to a large variety of trinucleotides and larger oligonucleotides, including single strands, double strands, and hairpin loops, by the techniques of equilibrium dialysis and nitrocellulose filter binding. I also propose to study the physical nature of the binding of oligonucleotides to IF3 by measuring ultraviolet light absorption and circular dichroism as a function of temperature, IF3 concentration and oligonucleotide concentration. In doing so, I expect to test the hypothesis I proposed in my preliminary work for the role of IF3 in protein synthesis. Finally, I propose to measure the kinetics of binding of fmet-tRNA met f to 30S ribosomal subunits and synthetic mRNAs of different secondary structures, in the presence of IF1, IF2, and GTP, while varying IF3 concentration.